I was wondering if there has been any research conducted on this issue. That is, all other things being equal how much more likely is a professed supporter of a political party or politician to support a position taken by that person or body, as opposed to the general likelihood of a person of that demographic. Specifically, how much does the political party's position influence that person, rather than how much more likely are they to affiliate to a party who already supports it...

As a hypothetical example, if we take say that an upper-middle class person (we could use the B-grade socio-economic status grouping if we were so minded) in a fairly cosmopolitan area, with a second degree level education, who spends around 25 hours per week on political research and say they might be 55% likely to support unlimited immigration. If we were to poll them and state that a political party/leader they specifically affiliate with opposes immigration, might we expect that to fall to 50%, 45%, 20%? Does any alteration in support vary by political affiliation, and by specific issue, or by political affiliation in concert with the specific issue concerned?

I'm looking for research with questions similar to the following combination:

  1. "Do you support [X]?"

  2. "[Your political party or leader here] supports [X]. Do you support [X]?"

Where [X] is the same political issue for both questions. The questions might be asked of the same person one after the other, or the questions might be asked of two randomised samples with the same professed party affiliation.

Has there been any research done on this subject?

  • 7
    Democrats are more likely to support a ban on Muslims if Trump's name is not mentioned, according to one poll: foxnews.com/politics/2015/12/18/…
    – Golden Cuy
    Commented May 18, 2016 at 12:17
  • 3
    "who spends around 25 hours per week on political research" = that seems to be an extreme edge case.
    – user1530
    Commented May 19, 2016 at 12:15
  • 2
    it was an example Commented May 19, 2016 at 12:15
  • 4
    I vaguely remember there being some sort of study done with #2, where X was changed to be the opposite of the what the leader actually supports. Many people "blindly" said they also support X. Unfortunately I can't remember where I heard/read this...
    – TTT
    Commented May 19, 2016 at 17:19
  • 4
    @TTT there was a recent 'experiment' done with Trump supporters I believe, they were given statements that Hitler originally made but for the purpose of the experiment some were attributed to Trump. Regardless of how te statement, the majority agreed with the sentiment when they thought their candidate made it
    – user7754
    Commented May 20, 2016 at 7:04

2 Answers 2



An article published in the Journal of Personality and Social Psychology (Cohen, 2003) conducted an experiment looking at something similar. What they found was that people agreed with policies that their party supported*. That wasn't very surprising, but they did find a few other things:

  • People agreed with their party even when their party's position was different than their personal opinion.
  • Their party's position was (causally) more important than the actual facts of the policy.
  • Participants reported that they evaluated policies based on the facts of the policy and their own political beliefs. Notably, this conflicts with every finding in the research.

As an example, a theoretical liberal and Democrat who supports welfare was likely to support a welfare policy which did not include food stamps, full health insurance, housing, daycare, tuition, and only $250 per month (more stringent than any real-world policy) - when they were told that the Democratic Party supported it.

This finding has been found many other times, such as this article and this one, both of which are in reputable political science journals.

* You asked "how likely is it..." and I answered "yes". Sorry about that. The research just doesn't lend itself to an answer in that format. The research doesn't frame it as a likelihood of an event ("what are the odds they agree with this position") but the extent of the agreement ("how much do they agree with this position?).


While not a scientific study, consider the 2016 US presidential election. A number of traditionally Democratic states shifted to the republican candidate, a change that is only amplified by the controversial nature of that candidate.

For better or for worse, a substantial number of Democratic voters didn't support the platform their party was espousing.

Based on that election, I would say that the working class Democrats voted their demographic, not their candidate.

  • 1
    Not sure this is clear. What does 'vote their demographic' mean? What demographic do working class democrats share with Donald Trump? It should also be noted that republican voter turnout increased in 2016, which likely had as much (if not, more) to do with the 'shift'.
    – user1530
    Commented Sep 17, 2017 at 21:11
  • 1
    This doesn't appear to answer the question. The question isn't about whether Democrats/Republicans voted for their own candidates. It's about the positions they support. Commented Sep 19, 2017 at 20:54
  • Their demographic was wanting a job. That should have been painfully clear in the last election - the candidate who campaigned on more jobs for US citizens and less immigrants to compete for those jobs... won most states where jobs were an issue, and gave him the electoral votes needed. Whether he will carry through with that promise, or even function as a president, is another matter. But, for better or for worse, that's why he won. A large number of traditionally democratic voters rejected the dem platform, which was heavy on PC and light on the need for current voters to stay employed.
    – tj1000
    Commented Sep 20, 2017 at 14:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .